The large increase in amount of data generated by digital systems yearns for more sophisticated approaches to data storing, processing, and analyzing. In this context, scale-out network-attached storage (NAS) systems have proved popular as a technology for managing this “Big Data.” Some scale-out NAS systems utilize a virtual Internet protocol (IP) failover scheme based on dynamic IP address allocation that distributes IP addresses to nodes participating in the IP address pool assigned to the NAS cluster. When allocating dynamic IP addresses within a clustered system, coordination of IP address ownership is critical to cluster operations. Moreover, the coordination is exclusive, i.e., only one network interface attached to a host (e.g., node of the NAS cluster) can own a network visible IP address at a time. A host can advertise ownership of an IP address by sending an Address Resolution Protocol (ARP) packet to a switch. Two or more network hosts attempting to claim the same IP address via ARP can lead to packet routing problems, known as an ARP conflict, on the network, resulting in the hosts becoming unreachable.
To avoid such ARP conflicts, IP management can be implemented, wherein a central IP manager can assign and revoke IP addresses exclusively to the hosts that are being managed. However, IP management via the single/central coordinator can have several drawbacks. For example, the coordinator is a single point of failure. In addition, the coordinator should be aware of each individual host's local state. As the size of the system grows, the amount of local state being replicated becomes untenable, and the likelihood of failure can significantly increase. These issues can cause problems in large distributed systems.
Instead of a central coordinator, a distributed IP management system can be implemented wherein IP addresses are dynamically procured by the hosts and assigned on a first-come, first-serve basis. However, in this scenario, a single host can be oversubscribed with IP addresses, resulting in an unbalanced distribution of IP addresses across a distributed system. This can lead to uneven resource usage and decrease the fault tolerance of the distributed system.
The above-described background relating to file systems is merely intended to provide a contextual overview of some current issues and is not intended to be exhaustive. Other contextual information may become further apparent upon review of the following detailed description.